ニューラルネット・遺伝的アルゴリズムを用いた粒子追跡画像計測法 Particle tracking velocimetry using neural networks and genetic algorithms
Particle tracking velocimetry (PTV) is one of the methods to measure flow velocity fields which is considered to be essential and useful for analyzing complex flow fields. This paper proposes a temporally particle pairing method using genetic algorithms and a flow velocity field estimation method using neural networks. The particle pairing method is based on spatial pattern relationship between pair-candidate particles with their respective neighbour particles in two exposures taken over a small time interval. If two particles are paired correctly, the spatial patterns of their pair-candidate particles are to be similar. The method finds correct pairs by applying a genetic algorithm. A potential problem of the method is that it can't measure velocity vectors at the points where no particles exist. The flow velocity field estimation method proposed in this paper solves it, which uses neural networks. The neural network is trained by using measured velocity vectors as teaching data so that the derivatives of a certain scholar function agree well with the measured data. The continuity equation of flow is consequently satisfied in the estimated vector fields and the scholar function gives the stream function.
- 可視化情報学会誌 = Journal of the Visualization Society of Japan
可視化情報学会誌 = Journal of the Visualization Society of Japan 18, 59-60, 1998-09-01
The Visualization Society of Japan